Panel - Hype vs. Reality: What’s Actually Working in AI Right Now?
Marko Aalto, Niina Hagman, Oguzhan Gencoglu, Pasi Helenius
AI is yielding results in personal productivity tools like ChatGPT and GitHub Copilot, but organizational applications still face challenges. Successful AI deployment requires strong leadership, clear business value, and integration into existing processes. Trust and transparency are crucial, especially in regulated industries. Overhyped benchmarks and lack of organizational readiness hinder progress. Emerging areas like synthetic data and improved human-computer interaction hold promise. The panelists emphasize the importance of measurement, continuous monitoring, and adapting to evolving AI capabilities.
"AI is only half of the coin; the other half is the data assets and systems you have. To get value from AI, the data part is crucial."
Summary
- AI is effective in personal productivity tools like ChatGPT but still has room for growth in organizational applications. - Successful AI projects require clear ownership, significant investment, and deprioritization of other initiatives. - Trust and transparency are crucial, especially in high-stakes industries; continuous monitoring and evaluation are necessary. - Generative AI benchmarks are often misleading; focus should be on tasks that can be fully automated. - Organizational skills and maturity are key to AI adoption; understanding AI's potential and aligning it with business needs is essential.
Article
AI reality check: experts separate hype from value at Helsinki tech forum
Leading industry figures discuss what's actually working in artificial intelligence and the barriers to wider implementation
The gap between AI's potential and its practical business value took center stage at the VERGE conference in Helsinki on Monday, as industry leaders gathered to separate genuine progress from inflated promises in artificial intelligence.
The panel discussion "Hype vs. Reality: What's Actually Working in AI Right Now?" brought together four AI experts – Marko Aalto, Niina Hagman, Oguzhan Gencoglu, and Pasi Helenius – who offered candid assessments of where artificial intelligence is delivering results and where it continues to fall short.
Personal productivity soars while businesses struggle
According to the panelists, AI has found its most successful applications in personal productivity tools. "On personal productivity and the use of AI assistants like ChatGPT or GitHub Copilot, those are definitely working and providing value. I use those every day, and I guess most of you do too," noted Marko Aalto, VP of Data and AI at Reactor.
However, the picture looks different at the organizational level. "On the business and organizational level there is still lots of room to discover, unfound and create value," Aalto added, highlighting that while individuals are benefiting from AI tools, companies are still searching for transformative applications.
Leadership gap holds back implementation
Niina Hagman, director of Data and AI Transformation Advisory at Dane Studios, pointed to leadership and resource allocation as critical factors in successful AI deployment. "It starts with passionate business owners who want to make a change. They're willing to invest their time, money, resources, and also willing to deprioritize something on the backlog," she explained.
The panel referenced a CIO survey showing that over 40% of failed AI projects trace back to weak cross-functional leadership or lack of ownership, underscoring that technical capability alone doesn't ensure success.
Benchmarks mislead while real value goes unmeasured
In a particularly revealing exchange, Oguzhan Gencoglu of Root Signals delivered a sharp critique of how AI progress is measured: "All these benchmarks of generative AI on really hard problems are complete noise. They give the wrong impression."
Gencoglu emphasized that what matters isn't incremental improvements in speed but complete automation. "What matters is I don't want to do all my tasks thirty percent faster. I want to completely automate thirty percent of my tasks end to end," he stated, suggesting that current evaluation methods don't capture real business impact.
Trust crucial in high-stakes industries
For Pasi Helenius, country manager of SaaS Finland, who works in highly regulated industries like finance and fraud protection, trust and transparency are non-negotiable. "In fraud, decisions must be made instantly for transactions to go through. Trust and transparency are crucial," he explained.
The panel agreed that continuous monitoring, evaluation, and clear audit trails become even more important as AI systems become more complex and non-deterministic.
Hidden opportunities: interfaces and synthetic data
Looking beyond the hype cycle, the panelists identified areas they believe offer untapped potential. Aalto suggested that user interfaces are due for reinvention: "The current visual interfaces are outdated. AI's understanding of humans opens new possibilities for human-computer interaction."
Meanwhile, Helenius highlighted synthetic data as "flying under the radar" despite showing early promise, particularly in sectors where data quality is paramount.
Human factors outweigh technical challenges
Perhaps the most consistent theme throughout the discussion was that organizational and human factors – not technology – present the biggest barriers to AI adoption.
"The tech is the easy part; it's the communication between humans that's the hard part," observed one panelist, with others nodding in agreement. Understanding business needs, securing stakeholder buy-in, and communicating results in human terms were repeatedly emphasized as more challenging than solving technical problems.
As the VERGE conference in Helsinki concluded, the message from these industry veterans was clear: while AI continues to transform how individuals work, organizations still face significant challenges in moving from experimentation to execution at scale. Success depends not just on choosing the right technology, but on leadership commitment, clear business value, and the human capacity to integrate AI into existing processes.
Part of VERGE | The AI Frontier: Creativity, Security & Collaboration